In a multi-person video conference setting, a video conferencing system will implement a process to focus the presentation on a particular individual, usually the person speaking at the time. For example, the conferencing system will detect a participant that is speaking and place the video feed associated in view for the other participants or give that participant's video feed a larger or more prominent space on the screen. Other participants (e.g., those not speaking or participating at the time) may be given a smaller portion of the screen for their respective video feeds, or not be shown at all.
The current approach of many video conferencing systems to focus only on a designated presenter or person speaking at the time is too basic and does not take into account that certain situations may benefit from other participants being emphasized or receiving a focus from the system. For example, a topic of conversation may relate to the expertise of a particular participant such that providing a focus on that person for others to view would be beneficial to the group. In a normal group conversation setting, where everyone is in the same room, people can generally look to certain people, whether they are speaking or not, for reactions and input during certain situations based on context. The present disclosure is directed to enhancing video conferencing systems in order to consider contextual information in determining how to present the video conference feed to participants in order to provide an experience which more closely resembles that of a face-to-face conference.